Closed T-Spline Surface Reconstruction from Medical Image Data

Abstract

The reconstruction from medical image data to CAD model is an essential process of Bio-medical engineering. So far, it is still challenging to create medical model which is suitable for both design and manufacturing. This paper introduces a process of constructing a single-patched closed T-spline surface model based on medical image data. The image data is first converted into triangular mesh and then parameterized onto a rectangular domain. An iterative least-squares fitting process is proposed to finally obtain the T-spline surface model with a user-specified tolerance. In this fitting process, the smoothing part of the algorithm is redesigned, considering the flexible structure of T-mesh and the geometric complexity of the T-spline surface. Local smoothness weighting factors are introduced to the fitting formula to locally adjust the smoothness of the surface. An adaptive smoothness checking points distribution method is proposed to reduce the computational cost. These algorithms are easy to implement and the obtained model is simple in form. Compared with the STL and NURBS model, the T-spline surface model requires less storage space and can be easily modified. The obtained model is suitable for Bio-medical engineering applications like bone scaffolds design, surgical planning and related manufacturing process.